Heterogeneously catalysed synthesis of new 3-alkoxypivalic acids from substituted 1,3-dioxanes

Article abstract of DOI:10.1006/jcat.2000.2956

Several methods for the preparation of the interesting and valuable compounds 3-alkyloxypivalic acids starting from substituted 1,3-dioxanes were examined. While one-step procedures combining isomerisation and oxidation were not successful due to degradation of formed intermediates and products, a two-step environmentally benign method could be established using O₂ as oxidant in the presence of heterogeneous catalysts. The scope of the method was tested by use of several 1,3-dioxanes and the isolated new pivalic acid derivatives were characterised.

Full text of DOI:10.1006/jcat.2000.2956

Journal of Catalysis 194, 294–300 (2000)
doi:10.1006/jcat.2000.2956, available online at http://www.idealibrary.com on
Heterogeneously Catalysed Synthesis of New 3-Alkoxypivalic
Acids from Substituted 1,3-Dioxanes
1
Jutta Fischer and Wolfgang F. H o¨ lderich
Department of Chemical Technology and Heterogeneous Catalysis, University of Technology RWTH Aachen, Aachen D-52074, Germany
Received January 28, 2000; revised May 2, 2000; accepted May 24, 2000
Additionally, when working in the presence of both H2
Several methods for the preparation of the interesting and valu- and ammonia, 3-alkyloxypropyl amines are formed via re-
able compounds 3-alkyloxypivalic acids starting from substituted ductive amination. These are valuable building units in or-
1
,3-dioxanes were examined. While one-step procedures combining
ganic synthesis (12).
isomerisation and oxidation were not successful due to degradation
of formed intermediates and products, a two-step environmentally
benign method could be established using O2 as oxidant in the pres-
ence of heterogeneous catalysts. The scope of the method was tested
by use of several 1,3-dioxanes and the isolated new pivalic acid
Seeing these results, it is surprising that there is no study
so far in the field of production of 3-alkyloxypropionic acids
which could also be valuable building units for organic syn-
thesis since they are containing a protected alcohol func-
tion. Furthermore, it is well known that the esters of pivalic
acid with sterically hindered alcohols give very stable syn-
thetic lubricants (13). The only 3-alkyloxypivalic acid men-
tioned in the literature is the 3-benzyloxypivalic acid, which
has been synthesised in several ways. It could be obtained
by reacting benzylic alcohol with pivalolactone or benzyl
bromomethyl ether with 2-lithio-2-methyl-propionic acid
derivatives were characterised.
�c 2000 Academic Press
Key Words: heterogeneous catalysis; O2 oxidation; 3-alkoxy-
pivalic acids.
1
. INTRODUCTION
The isomerisation of1,3-dioxanesto 3-alkyloxypropanals (14, 15). Also, oxidation of 3-benzyloxypivaldehyde could
is known to occur in the presence of various heterogeneous be achieved with potassium permanganate or with oxygen
catalysts (1–7). The catalysts include silica impregnated catalysed by salts of cobalt or manganese (2, 16). The tert-
with oxides or hydroxides of aluminum or lanthanides, butylhydroperoxide ester of this acid has been found to be
acid phosphates having a zeolite structure, and acid-treated an effective polymerisation initiator for PVC (2).
metal oxides. Of main importance however are zeolites of
However, these synthetic methods are environmentally
the pentasil family which are highly active and selective in not benign because of the high amount of inorganic salts
the isomerisation reaction (5). By employing 5,5-dimethyl- formed during the reactions, the use of bromine-containing
substituted 1,3-dioxanes in the isomerisation reaction, pi- compounds, and the stochiometric character of the oxida-
valic aldehyde derivatives are formed. The mechanism tion with permanganate (17). Besides, the high price of
of this interesting reaction is discussed by Rondestvedt the starting materials makes the reactions economically un-
(
6).
favourable.
By introducing hydrogen into the reaction system and
Being interested in new substances having a neostructure
modifying the catalyst by impregnation with copper or in the position � to the functional group, it was our aim
platinum, the corresponding 3-alkyloxypropanols can be to find an easy and generally applicable method for the
obtained (8–10). These are not formed via a consecutive synthesis of 3-alkyloxypivalic acids.
reaction pathway from the aldehydes but by direct hydro-
genation of the 1,3-dioxanes (7, 11).
2
. METHODS
The neoalkanals as well as the neoalcohols found use
in organic synthesis for the preparation of herbicides and
biologically active chemicals. Also, the phthalates of the al-
cohols are effective plasticisers for PVC, showing less ten-
dency for migrating than dioctylphthalate (DOP).
2
.1. Reaction Conditions
One-step batch experiments were carried out in a steel
autoclave with a volume of 70 ml tested for pressures up
to 100 bar equipped with a glass-coated magnetic stirrer
and two high-pressure valves. Starting materials (0.1 g of
catalyst, 1 g of 1,3-dioxane, 2 ml of benzene) were placed in
1
To whom correspondence should be addressed. Fax: 0049-241-
8
888291. E-mail: hoelderich@rwth-aachen.de.
294
0021-9517/00 $35.00
Copyright �c 2000 by Academic Press
All rights of reproduction in any form reserved.

NEW 3-ALKOXYPIVALIC ACIDS FROM 1,3-DIOXANES
295
the autoclave; after the autoclave was closed, oxygen was equipped with a magnetic stirrer. By use of a heating bath,
added up to the desired pressure (2 to 30 bar) with stirring. 1 l of the reaction mixture from the isomerisation scale-up
The autoclave was heated to the desired temperature (250 was brought to the selected temperature and a heteroge-
�
to 300 C) using a heating band and a temperature controller neousoxidation catalyst wasadded. Oxygen wasintroduced
(
Eurotherm comp. 91e). Reaction time varied between 1 into the stirred reaction mixture via the tube for 6 h; the oxy-
and 24 h. Since the carboxylic acids formed are not volatile gen flow was then stopped, and the mixture was allowed to
and thus cannot be analysed by GC, they were transferred cool and the catalyst to settle. Samples were treated with
into the methyl esters by adding diazomethane (CH2N2) CH2N2.
until N2 formation stopped.
In the case of the discontinuous two-step reactions, the 2.2. Starting Materials and Catalysts
procedure was carried out employing the reaction param-
eters described above without the presence of oxygen. O2
1
,3-Dioxanes were prepared by reaction of neopentyl-
glycol (NPG) with the appropriate aldehydes catalysed by
Amberlyst 15. Equimolar amounts of aldehyde and NPG
were placed into an one-neck round-bottom flask equipped
with a Dean-Stark apparatus. Three hundred milliliters of
petroleum ether and 1 g of Amberlyst 15 were added per
mole of aldehyde. The vessel was placed in an oil bath and
(
20bar) wasadded after coolingfollowingthe isomerisation
�
step; the mixture was then heated to 50 C for 6 h. Samples
were prepared for GC analysis by adding CH2N2.
Continuous gas-phase experiments were carried out un-
der atmospheric pressure. Two grams of catalyst pellets was
held in a fixed bed by a metal sieve at the end of a tube re-
actor. A solution of 1 g of 1,3-dioxane in 5 g of toluene was
introduced via a pump into the evaporating zone where
the carrier gas entered the reactor. The temperature was
�
heated to 90 C under stirring. After the theoretically pre-
dicted amount of water had been formed, petroleum ether
was removed under atmospheric pressure. Products were
isolated by distillation under reduced pressure. Starting ma-
terials as well as solvents were obtained from Fluka and
used without further purification.
Boron-containing pentasil zeolites (B-MFI, catalyst A)
were kindly provided by BASF, and Al-MFI-catalysts (cata-
lyst B) from Degussa. Charcoal (catalyst F) was bought
from Aldrich. Synthesis of V-MCM-41 (catalyst C) as well
as impregnation and other treatments were carried out in
our laboratories.
�
varied between 200 and 350 C, and the amount of carrier
gas was 20 l/h with the N2/O2 ratio varying from pure oxy-
gen to pure nitrogen. The reactor was placed in a furnace
where it could be heated from two sides; isothermic reac-
tion conditions were achieved by a motor-driven ventilator.
The weight hourly space velocity (WHSV) was in a range
�
1
from 0.25 to 1.5 h . The reaction mixture was collected in
a double-walled glass flask which was equipped with a re-
�
flux condenser cooled to � 15 C. Reactions were run for 6 h
For impregnation of the acid catalysts, 10 g of catalyst
powder was stirred in 50 ml of water and the following
amounts of metal salts were added:
with samples taken every 2 h.
In the two-step continuous reaction, a tube was led
through the condenser into the flask which was heated using
its double wall. Pure nitrogen (12 l/h) was used as a carrier
gas for the isomerisation reaction at atmospheric pressure
Ag O
0.159 g of silver nitrate
2
�
and within a temperature range from 200 to 350 C. Other
(giving catalyst D using B-MFI)
reaction conditions are the same as those described above.
Oxygen (20 l/h) was introduced via the tube, dispersing a
heterogeneous oxidation catalyst in the collected reaction
V2O5
0.263 g of ammoniumtetravanadate
(
giving catalyst E using B-MFI).
mixture. To accommodate the higher gas stream, the tem-
�
perature of the condenser was lowered to � 30 C. Reactions The mixture was stirred for 30 min at room temperature
�
were run for 6 h. Samples were modified by CH2N2.
For a scale-up of the two-step continuous reaction, a pilot calcined at 550 C to form the oxides.
and dried at 60 C at reduced pressure. The material was
�
plant wasused. The reactor (dimensions:length 80cm, inner
For the synthesis of catalyst C, a gel of the following com-
diameter 1 cm) was fitted with a metal sieve at the bottom to position was used: 4.40 g of SiO2, 0.26 g of VOSO4 �5H2O,
hold the catalyst fixed bed. To prevent loss of catalyst, 60 ml 22.04 g of H2O, 3.78 g of tetradecyl-trimethyl-ammonium
of glass beads (diameter 3 mm) was placed on the sieve; a bromide (TDTMABr), and 11.70 g of a 20% solution
7
cm layer of the zeolite catalyst was added. To allow better of tetraethylammonium hydroxide (TEAOH). In a poly-
evaporation of the reaction feed, it was covered with 25 ml propylene flask, 18% of the SiO2 was suspended in water.
of glass beads (diameter 0.6 mm). The temperature was TDTMABr, TEAOH, and VOSO4 were added in that order
measured both at the outside of the reactor at the level of with stirring. After the suspension was stirred for 1 h, the
the catalyst bed and within the catalyst bed itself.
remaining silica was added and the gel stirred for 5 h. After
�
For a scale-up of the oxidation, a tube was led through 7 days at 100 C the product was filtered, washed, dried at
�
�
�
a reflux condenser (� 30 C) into a 1 l round-bottom flask 100 C, and calcined in air for 6 h at 550 C.

¨
FISCHER AND HOLDERICH
2
96
TABLE 1
and oxidation together in one reactor by introducing oxy-
gen into the reaction system, allowing the formed aldehyde
to oxidise to the corresponding acid. This strategy was tried
Properties of the Isomerisation Catalysts
Metal
BET
Micropore
volume
Mesopore both in batch reactions and in continuous gas-phase reac-
Si : M(III) content surface
Catalyst (mol/mol) (wt% ) (m g
volume
tions. As a test molecule we used 2-phenyl-5,5-dimethyl-
,3-dioxane (1). The conversion according to Eq. [1] gives
2
� 1
3
� 1
3
� 1
)
(cm g
)
(cm g )
1
the known 3-benzyloxypivalic acid (3) as a product with
3-benzyloxypivalic aldehyde (2) as an intermediate.
A
B
C
D
E
16.3
13.3
No M(III)
18.0
None
None
1
1
1
358
0.133
0.119
No micropores
0.129
0.29
0.06
0.80
0.22
0.25
353
1005
351
17.2
344
0.127
Note. Element composition determined by ICP-AES; surface and pore
volume measured by N2 adsorption.
The stability of the starting material and products was ex-
As oxidation catalysts, several silver-containing materials pected to be a problem since isomerisation runs effectively
�
were used: pure silver oxide (Ag2O, catalyst G) obtained only at temperatures higher than 250 C. At this tempera-
from Fluka, and an Al2O3 catalyst loaded with 50% silver ture, decarbonylation of radical species formed during the
oxide (Ag2O, catalyst H), kindly provided by BASF.
oxidation of aldehydes with oxygen is strongly favoured,
Silver on charcoal (catalyst I) was prepared by impreg- especially in the case of 2,2-substituted aldehydes since the
nation of charcoal with Ag2CO3. A 0.6 g portion of silver substituents can stabilise the resulting alkyl radical. Fur-
carbonate was dissolved in 400 ml of water and the solution thermore, the ether function in the isomerisation products
�
brought to 90 C. A 2.2 g amount of charcoal was added and offers an additional point for the attack of oxygen or the
the mixture was stirred for 16 h. Water was removed under acid catalyst.
�
reduced pressure at 60 C. The material was calcined under
a nitrogen flow to avoid CO2 formation from charcoal. The degradation processes either by working under pressure to
Our aim was therefore to reduce the extent of these
�
�
temperature was first raised to 200 C at a rate of 3 C/min, suppress further gas formation (batch reactions) or by us-
held for 1 h, and then raised to 300 C at a rate of 2 C/min ing short contact times with the catalyst to avoid cleavage
�
�
and held for 5 h.
The properties of the isomerization catalysts are summa-
rized in Table 1.
of products at the acid catalyst (continuous reactions).
It was found that in batch reactions formation of ben-
zoic acid (4) and its esters was the main process. Benzoic
acid can be formed by oxidation either of 1, of different iso-
merisation products, or of benzaldehyde (5) which results
from cleavage of 1 or isomerisation products at the acid
catalyst due to the long contact time. Varying the reaction
parameters (reaction time, pressure, oxygen content) did
not increase the yield of the desired neoacid.
2
.3. Reaction Products
Yields were determined by GC analysis. Reaction prod-
ucts were intensively characterised by GC–MS and after
1
isolation by IR and NMR spectroscopy ( H NMR and
13
C NMR). Isolation of the new 3-alkoxypivalic acids was
These results illustrated clearly the necessity to work
in the gas phase with a high weight hourly space velocity
achieved by selective extraction. Acidic compounds were
isolated by washing the reaction mixture with sodium bi-
carbonate solution. The aqueous solution was washed with
diethylether and acidified with HCl. The separating solid
or oil was extracted with diethylether, which was dried with
Na2SO4 and evaporated.
(
WHSV) to ensure a short contact time at the catalyst.
Boron-containing pentasil zeolites (catalyst A) which
are known as the best catalysts under these conditions (5)
were impregnated with metal oxides known as oxidation
Except for 3-benzyloxypivalic acid these are new sub-
stances not previously described in the literature. Their
chemical and toxicological properties will be examined
shortly after a scale-up of their production has been car-
ried out.
TABLE 2
Comparison of Different Catalysts in the Isomerisation
of 1 under Oxidising Atmosphere
Conversion Selectivity Selectivity Selectivity Selectivity
Catalyst
of 1 (% )
to 2 (% )
to 3 (% )
to 5 (% )
to 4 (% )
3
. RESULTS AND DISCUSSION
A
D
E
69
51
54
23
65
46
6
0
2
1
0
6
10
17
69
4
6
4
3
.1. Isomerisation of 1,3-Dioxanes
in an Oxygen Atmosphere
C
100
17
In analogy with the synthesis of neoalcohols and
Conditions. 290 C; 15 l/h O2; WHSV = 1 h� 1; TOS = 6 h.
�
neoamines, our first strategy was to conduct isomerisation

NEW 3-ALKOXYPIVALIC ACIDS FROM 1,3-DIOXANES
297
TABLE 3
Influence of Temperature and WHSV with Catalyst D
Conversion
Selectivity
Selectivity
Selectivity
Selectivity
Reaction conditions
of 1 (% )
to 2 (% )
to 3 (% )
to 5 (% )
to 4 (% )
�
�
�
� 1
2
2
2
90 C, WHSV = 1 h
51
52
37
65
54
62
2
6
6
10
6
3
6
4
2
� 1
� 1
90 C, WHSV = 0.5 h
50 C, WHSV = 0.5 h
Conditions. 15 l/h O2; TOS = 6 h.
catalysts, such as V2O5 and Ag2O. Resulting catalysts were was a further reduction of the WHSV. This, however, led
Ag-B-MFI (D) and V-B-MFI (E). In addition to these, a to a highly explosive mixture making further investigation
V-MCM-41 material (C) was used which contained vana- dangerous and because of necessary safety measures eco-
dium in lattice positions. Its larger pore size should pose no nomically unfavourable.
constraints for both starting materials and reaction prod-
ucts. Also, it was hoped that isolated V-species might allow tionally, it was observed that the ether function-containing
a more selective oxidation. compounds underwent fast oxidation and degradation in
Therefore, this approach was no longer followed. Addi-
As shown in Table 2, the selectivity to 2 obtained by cata- the presence of oxygen even at temperatures as low as
�
lyst A after 6 h time on stream (TOS) is very low, not ex- 120 C. As a consequence, a two-step procedure had to be
pected from published results (5). The presence of oxygen employed, separating the high-temperature isomerisation
has to be a limiting factor. The metal oxide impregnated from the oxidation procedure.
catalysts D and E show higher selectivities to the aldehyde,
but form only 1 to 2% of the corresponding acid. The lower
3
.2. Two-Step Reaction for the Synthesis
of 3-Benzyloxypivalic Acid
conversions are due to blocking of some acidic centres by
metal oxide. In TPD experiments it was found that by silver
oxide impregnation the number of acid sites is reduced by
In two-step batch experiments, isomerisation catalyst A
1
5% . Thus, a somewhat lower activity in the isomerisation was studied by variation of several reaction parameters. The
reaction is to be expected. Vanadium especially seems to focus lay on the influence of the reaction time since avoid-
favour the formation of benzaldehyde, either because its ing the formation of a higher amount of by-products was
oxophilic character makes desorption more difficult, thus the main target. Selectivity to the aldehyde was observed
allowing consecutive reactions, or by facilitating an elec- to drop slowly after 16 h while formation of benzaldehyde
tron shift in the starting material by its ease in changing increases.
oxidation states.
In the oxidation step, silver oxide (catalyst G) was used
In the case of catalyst D some acid was obtained. There- for optimising and comparison, because it is known as a
fore, this catalyst was studied further. The low fraction of suitable catalyst for the oxidation of aldehydes (18). The
aldehyde oxidised to the acid could be the result of too combined influence of silver oxide and of the isomerisation
short contact time at the catalyst. Furthermore, the tem- catalyst was studied. While catalyst G does slightly catalyse
perature was reduced to decrease consecutive oxidation of the reaction (Table 4), even without a catalyst the aldehyde
products.
is nearly completely oxidised. Selectivity is even higher in
As shown in Table 3, a lower WHSV indeed leads to an the uncatalysed reaction. The reason is the high concentra-
increase of selectivity to 3, while a lower temperature only tion of O2 in the liquid phase due to the pressure of 20 bar,
decreases conversion of 1,3-dioxane. The obvious next step resulting in uncatalysed autoxidation of the aldehyde. The
TABLE 4
Influence of Isomerisation and Oxidation Catalysts on the Oxidation of 2
Conversion
Selectivity
Selectivity
Selectivity
Selectivity
Reaction conditions
of 1 (% )
to 2 (% )
to 3 (% )
to 5 (% )
to 4 (% )
No filtration, no catalyst G
No filtration, catalyst G
Filtration, catalyst G
81
85
70
2
0
0
53
50
48
6
8
8
5
6
3
�
�
Conditions. Isomerisation at 300 C; batch reaction for 16 h; catalyst A; oxidation at 50 C; 20 bar O2;
batch reaction for 6 h; m(dioxane)/m(catalyst) = 10 for both reactions.

¨
FISCHER AND HOLDERICH
2
98
TABLE 5
Comparison of Different MFI Catalysts in the Two-Step Synthesis
of 3-Benzyloxypivalic Acid
Conversion Selectivity Selectivity Selectivity Selectivity
Catalyst
of 1 (% )
to 2 (% )
to 3 (% )
to 5 (% )
to 4 (% )
A
B
70
97
0
2
48
27
8
18
3
4
�
�
Conditions. Isomerisation at 300 C; 16 h; oxidation at 50 C; 20 bar O2;
h; catalyst G; m(dioxane)/m(catalyst) = 10 for both reactions.
6
FIG. 1. Influence of temperature on the oxidation of 2 with oxygen.
Oxidation time, 6 h; 0.05 g catalyst G/12 g 1.
presence of the isomerisation catalyst does not impede ox-
idation.
By a combination of isomerisation and oxidation steps,
the yield of 3 could be optimised. Substantial amounts up to
a selectivity of 48% were obtained (Table 5). The lower se-
lectivity and higher conversion in the case of catalyst B are
due to the higher acidity compared to A. Stronger acid cen-
tres favour cleavage of reaction products to benzaldehyde
by slowing down desorption of 2 or 3.
�
� 1
found to be about 300 C and a WHSV of 1 h ; both are in
accordance with data given in the literature (7).
For the oxidation step, after determining the optimum
temperature using catalyst G as a model catalyst, the influ-
ence of different oxidation catalysts was studied (Table 6).
Next to several silver-containing materials, charcoal (F) was
tested as a catalyst since it has been described in the litera-
ture to accelerate aldehyde oxidation due to itslarge surface
area (19).
Using silver oxide as a catalyst (G), conversion can be
slightly increased. Charcoal (F) does not give satisfactory
results, while its impregnation with silver (I) leads to a very
high selectivity at nearly complete conversion. With sil-
ver oxide-impregnated Al2O3 (H), selectivity is somewhat
lower, probably due to the higher acidity of the carrier.
While selectivity drops slightly with an increase of tem-
Due to the low volatility of 3-alkyloxypivalic acids, prod-
ucts were isolated by use of a selective extraction procedure
(
see Section 2.3).
By this procedure, all acid components of the reaction
mixture are isolated. Therefore, any formation of benzalde-
hyde during the isomerisation step and consequently of
benzoic acid in the oxidation reaction leads to contami-
nation of 3 by 4. As long contact times are inherent to the
batch isomerisation and thus necessarily lead to benzalde-
hyde by consecutive reactions, a semicontinuous combina-
tion of gas-phase isomerisation and liquid-phase oxidation
was developed (see Section 2). Thereby, the temperature
for the oxidation step could be reduced drastically, allow-
ing the substitution of benzene by the more environmen-
tally benign toluene.
To find optimal reaction conditions for the transfer of dif-
ferent 1,3-dioxanes as starting materials, reaction parame-
terswere varied in both the isomerisation and the oxidation.
Of primary importance for running the isomerisation in
an economically favoured region are the reaction temper-
ature and the WHSV. Optimal reaction conditions were
�
perature (Fig. 1), conversion is nearly complete at 80 C.
Higher temperatures are problematic because of the close-
ness to the boiling point of toluene used as a solvent.
As a cheaper alternative to O2, air was tested as an oxi-
dant. As expected, conversion was by far lower than that in
the case of pure oxygen (57% at 80 C) although the total
amount of gas was doubled. Surprisingly, selectivity to the
�
�
neoacid increases with conversion, going from 23% at 50 C
TABLE 6
Influence of Oxidation Catalysts in the Two-Step Synthesis of 3
Catalyst
Conversion of 2 (% )
Selectivity to 3 (% )
Without
90
95
96
91
62
76
65
80
89
71
G
H
I
FIG. 2. Scale-up of the isomerisation of 1 (factor 5). Temperature,
F
�
� 1
3
00 C; carrier gas, N2, 12 l/h; WHSV = 1 h ; atmospheric pressure;
�
Conditions. 80 C; O2 20 l/h; m(dioxane)/m(catalyst) = 1200.
catalyst A.

NEW 3-ALKOXYPIVALIC ACIDS FROM 1,3-DIOXANES
299
TABLE 7
TABLE 8
Scale-up of the Oxidation
Isomerisation of Different 1,3-Dioxanes
Scale
Conversion of 2 (% )
Selectivity to 3 (% )
Starting
material
Conversion of
1,3-dioxane (% )
Selectivity to 3-alkoxypivalic
aldehyde (% )
Laboratory (70 ml)
Scale-up (1 l)
95
86
65
84
6
7
8
9
58
62
60
41
41
93
92
85
91
85
�
Conditions. 80 C; O2 20 l/h; m(dioxane)/m(catalyst) = 1200; catalyst G.
10
�
�
� 1
to 65% at 80 C. Main by-products are formed by decar-
Conditions. 300 C; carrier gas, N2 12 l/h; WHSV = 1 h ; atmospheric
pressure; catalyst A.
bonylation. Since it is known that the oxidation of aldehy-
des to carboxylic acids is slightly hindered by water, the fact
that the air was not dried prior to use might have favoured
decarbonylation which can occur at lower temperatures.
3
.4. Synthesis of New 3-Alkyloxypivalic Acids
The route described above should be generally applica-
ble for the synthesis of new 3-alkoxypivalic acids. For this
reason different 1,3-dioxanes have been converted to acids.
3
.3. Scale-up of the Two-Step Reaction
Scale-up of the reaction was performed in a pilot plant for The chosen reaction conditions were those optimised for
the isomerisation reaction. In the same step the duration of the case of 1.
the catalyst activity was tested. The reaction was run for 50 h
The 1,3-dioxanes synthesised and used as starting mate-
time on stream (TOS) during which a decrease in catalyst rials in the reactions are shown in Fig. 3.
activity was observed.
In the isomerisation reaction using catalyst A, all diox-
Deactivation sets in immediately at the start of the re- anes show high selectivities (Table 8); the lower conversions
action but proceeds more slowly after a TOS of 20 h, con- in the cases of 9 and 10 are probably due to steric hindrance
version being almost stable at about 45% during the second of the two-point adsorption of the dioxanes at the catalyst
half of the reaction time. The selectivity is very high at about surface necessary for the isomerisation mechanism.
9
0% and does not change significantly during the reaction
The neoaldehydes show high sensitivity to degradation
by oxygen. Figure 4 shows the results for the two-step reac-
(
see Fig. 2).
In oxidation, a scale-up leads to a decrease in aldehyde tion on a laboratory scale. While all starting materials give
conversion. The reason for this is the lower amount of oxy- satisfactory results, the loss of selectivity compared to the
gen brought into the reaction mixture, since an increase of isomerisation step is mainly due to the formation of car-
the gas stream in scale with the increase of reaction mix- boxylic acids corresponding to the aldehydes used in the
ture would have overcome the cooling capacity of the reflux synthesis of the dioxanes.
condenser. On the other hand, a higher selectivity to 3 is
observed, thus keeping the overall yield in the same region (see Section 2). Except for 3-(2-methylphenyl-methoxy)-
Table 7).
pivalic acid, the obtained neoacids were oils. They were
All acids could be isolated via the described method
(
FIG. 3. 1,3-Dioxanes used as starting materials in the two-step synthesis of 3-alkyloxypivalic acids.

¨
FISCHER AND HOLDERICH
3
00
This way, pivalic acid derivatives containing propoxy, isobu-
toxy, and n-butoxy, 2-ethyl-butoxy, and 2-methylbenzyloxy
groups could be obtained.
ACKNOWLEDGMENTS
The authors are very grateful to “Deutsche Forschungsgemeinschaft”
(
German Science Foundation) for the financial support within the scope
of the “Sonderforschungsbereich 442.”
REFERENCES
FIG. 4. Conversion of different 1,3-dioxanes and selectivity to the
1. Rondestvedt, C. S., and Mantell, G. J., J. Am. Chem. Soc. 82, 6419
(1960).
2. Arpe, H. J., DOS 29 22 698, Dec. 11, 1980 (Hoechst AG).
3. H o¨ lderich, W. F., and Merger, F., EP 0 291 807 B1, Jan. 15,
�
corresponding 3-alkyloxypivalic acids. Temperature, 300 C; carrier gas,
�
1
N2, 12 l/h; WHSV = 1 h ; TOS = 6 h; atmospheric pressure; oxidation at
�
8
0 C; O2, 20 l/h; m(dioxane)/m(catalyst) = 1200; catalyst G.
1992 (BASF AG).
4
5
. H o¨ lderich, W. F., and Merger, F., EP 0 291 806 B1, Aug. 26, 1992 (BASF
AG).
. H o¨ lderich, W. F., Merger, F., and Fischer, R., EP 199 210, Oct. 29, 1988
characterised spectroscopically and mass spectrometrically
as described in the Experimental section.
(
BASF AG).
6
7
. Rondestvedt, C. S., J. Am. Chem. Soc. 84, 3319 (1962).
. Paczkowski, M. E., and H o¨ lderich, W. F., Stud. Surf. Sci. Catal. 83, 399
4
. CONCLUSIONS
(
1994).
8. Paczkowski, M. E., and H o¨ lderich, W. F., J. Mol. Catal. A: Chem.
18(3), 311 (1997).
. Paczkowski, M. E., and H o¨ lderich, W. F., J. Mol. Catal. A: Chem.
18(3), 321 (1997).
0. H o¨ lderich, W. F., Paczkowski, M., Heinz, D., and Kaiser, Th., DE 196
1703, May 30, 1996 (Hoechst AG-Ruhrchemie).
An easy and generally applicable method for the synthe-
sis of 3-alkyloxypivalic acids has been developed and five
new substances of this class could be isolated and charac-
terised.
1
9
1
1
While a one-step procedure performing the isomerisa-
2
tion of 1,3-dioxanes in an oxygen atmosphere was un- 11. Bart o´ k, M., and Apjok, J., Acta Phys. Chem. 21(1–2), 69 (1975).
1
2. H o¨ lderich, W. F., Paczkowski, M., and Heinz, D., DE 196 21704, May
0, 1996 (Hoechst AG-Ruhrchemie).
successful due to degradation of starting materials and
products, a two-step process was developed using 2-phenyl-
3
1
3. Weigert, W. M., in “Ullmanns Enzyklop a¨ die der Industriellen
Chemie” (E. Bartholom e´ , E. Biekert, H. Hellmann, and H. Ley, Eds.),
Vol. 9, 4th Ed., p. 144. Verlag Chemie, Weinheim, 1975.
5
,5-dimethyl-1,3-dioxane as a test material.
Combining a continuous gas-phase isomerisation with a
discontinuous oxidation reaction leads to good yields of 3- 14. Luhmann, U., and L u¨ ttke, W., Chem. Ber. 105, 1350 (1972).
1
1
5. Matsushita, Y.-I., Chem. Lett. 1397 1983).
6. Abiko, A., Roberts, J. C., Takemasa, T., and Masamune, S., Tetrahedron
Lett. 27(38), 4537 (1986).
7. H o¨ lderich, W. F., in “Proceedings, 10th International Congress on
Catalysis, Budapest, 1992” (L. Guczi, F. Solymosi, and P. Tetenyi, Eds.),
p. 127. Akademia Kiado, Budapest, 1992; Stud. Surf. Sci. Catal. 78, 127
benzyloxypivalic acid. High selectivities up to 60% could
be obtained and the acid could be isolated in 98% purity.
A scale-up of the reaction could be achieved with good
results in both the isomerisation and the oxidation steps.
While isomerisation led to a selectivity of more than 90% at
a conversion of 45% , the formed 3-benzyloxypivalic alde-
hyde could be oxidised to 86% with a selectivity of 84%
giving an overall yield of 30% in the scale-up reaction.
Transfer to several different starting materials showed
the broad range of application of the developed method.
1
(
1993).
1
1
8. Waters, W. A., in “Mechanisms of the Oxidation of Organic Com-
pounds” (H. J. Emel e´ us, D. W. G. Style, and R. P. Bell, Eds.). Academic
Press, London, 1964.
9. Henecka, H., in “Methoden der Organischen Chemie/Houben-Weyl”
(E. M u¨ ller, Ed.). Georg Thieme Verlag, Stuttgart, 1952.